The Impact of Southern Ocean Residual Upwelling on Atmospheric CO[Subscript 2] on Centennial and Millennial Timescales
Total Page:16
File Type:pdf, Size:1020Kb
The impact of Southern Ocean residual upwelling on atmospheric CO[subscript 2] on centennial and millennial timescales The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters. Citation Lauderdale, Jonathan M., Richard G. Williams, David R. Munday, and David P. Marshall. “The Impact of Southern Ocean Residual Upwelling on Atmospheric CO2 on Centennial and Millennial Timescales.” Climate Dynamics (May 10, As Published http://dx.doi.org/10.1007/s00382-016-3163-y Publisher Springer Berlin Heidelberg Version Final published version Citable link http://hdl.handle.net/1721.1/107158 Terms of Use Creative Commons Attribution Detailed Terms http://creativecommons.org/licenses/by/4.0/ Clim Dyn DOI 10.1007/s00382-016-3163-y The impact of Southern Ocean residual upwelling on atmospheric CO2 on centennial and millennial timescales Jonathan M. Lauderdale1 · Richard G. Williams2 · David R. Munday3 · David P. Marshall4 Received: 27 January 2016 / Accepted: 29 April 2016 © The Author(s) 2016. This article is published with open access at Springerlink.com 6 3 1 Abstract The Southern Ocean plays a pivotal role in cli- a rate of typically 1–1.5 parts per million/10 m s− by mate change by exchanging heat and carbon, and provides enhancing the communication between the atmosphere and the primary window for the global deep ocean to commu- deep ocean. This metric can be used to interpret the long- nicate with the atmosphere. There has been a widespread term effect of Southern Ocean dynamics on the natural car- focus on explaining atmospheric CO2 changes in terms bon cycle and atmospheric CO2, alongside other metrics, of changes in wind forcing in the Southern Ocean. Here, such as involving the proportion of preformed nutrients and we develop a dynamically-motivated metric, the residual the extent of sea ice cover. upwelling, that measures the primary effect of Southern Ocean dynamics on atmospheric CO2 on centennial to Keywords Southern Ocean · Atmospheric carbon dioxide · millennial timescales by determining the communication Ocean overturning · Ocean residual circulation · Ocean with the deep ocean. The metric encapsulates the com- upwelling · Ocean carbon cycle · Climate metric bined, net effect of winds and air–sea buoyancy forcing on both the upper and lower overturning cells, which have been invoked as explaining atmospheric CO2 changes for 1 Introduction the present day and glacial-interglacial changes. The skill of the metric is assessed by employing suites of idealized The Southern Ocean plays a fundamental role in the cli- ocean model experiments, including parameterized and mate system, ventilating much of the global ocean by form- explicitly simulated eddies, with online biogeochemistry ing subtropical mode, intermediate and bottom waters, as and integrated for 10,000 years to equilibrium. Increased well as uniquely returning deep waters to the surface (Mar- residual upwelling drives elevated atmospheric CO2 at shall and Speer 2012). By providing the window for the global deep ocean to communicate with the atmosphere, these different physical transport pathways lead to a deli- * Jonathan M. Lauderdale cate carbon balance over the Southern Ocean (Morrison [email protected] et al. 2015): atmospheric CO2 can be drawn down by sub- * Richard G. Williams duction of upper ocean waters, as well as increased by out- [email protected] gassing of CO2 from natural carbon-rich waters. Changes in physical forcing over the Southern Ocean 1 Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA have been invoked as explaining changes in atmospheric 02139‑4307, USA CO2 for the past (Sigman and Boyle 2001; Toggweiler 2 Department of Earth, Ocean and Ecological Sciences, et al. 2006; Menviel et al. 2008; Tschumi et al. 2008, 2011; School of Environmental Science, University of Liverpool, d’Orgeville et al. 2010; Sigman et al. 2010) and the pre- Liverpool L69 3GP, UK sent day (Frölicher et al. 2015; Morrison et al. 2015; Zick- 3 British Antarctic Survey, Cambridge CB3 0ET, UK feld et al. 2007). For example, stronger winds have been 4 Department of Physics, University of Oxford, Oxford OX1 invoked in the present day as driving a weakening in the 3PU, UK sink of anthropogenic carbon, inducing enhanced surface 1 3 J. M. Lauderdale et al. mixing (Le Quéré et al. 2007), increased upwelling (Len- to millennial timescales. Here our focus is on changes in ton and Matear 2007) and a stronger Eulerian-mean over- the natural carbon cycle, so that ocean carbon inventory turning (Lovenduski et al. 2007, 2008), which increases changes equal opposing atmospheric CO2 changes. The the exposure of natural carbon-rich deep waters to the sur- residual upwelling is defined by the upwelling over the face, opposing anthropogenic carbon uptake. Changes in Southern Ocean, including the effects of both the upper and the zonal symmetry of Southern Hemisphere winds have lower overturning cells (Fig. 1), accounting for the com- recently been implicated in the reinvigoration of the South- bined effects of the Eulerian-mean and eddy circulations. ern Ocean carbon sink (Landschützer et al. 2015). Stronger The upper cell consists of Upper Circumpolar Deep Water winds have also been proposed as strengthening upwelling drawn to the surface, which is then transported northward and initiating the deglacial rise in atmospheric CO2 (Ander- across the ACC and then subducted as Antarctic Interme- son et al. 2009). However, the paleoclimatic inferences of diate and Subantarctic Mode Water (Talley 2013). This implied upwelling have often been interpreted in terms of transformation to mode waters requires a buoyancy input wind-induced changes in the Eulerian-mean circulation, from heating or freshwater fluxes to lighten the upwelled rather than accounting for the partial compensation from waters. The lower cell involves upwelling confined south the ocean eddy circulation. of the ACC, transferring Lower Circumpolar Deep Water to There is a debate over the expected response of the the surface, which then becomes denser and contributes to Southern Ocean circulation to altered wind forcing. From the formation of bottom waters adjacent to Antarctica (Tal- diagnostics of high-resolution numerical models, strength- ley 2013). This transformation of surface to bottom waters ening winds may lead to little overall increase in strength requires a buoyancy loss from cooling and brine rejection, of the large-scale Southern Ocean horizontal circulation, which may be sensitive to the strength of the polar easterly usually inferred from zonal Antarctic Circumpolar Current winds (Stewart and Thompson 2012). Our metric of resid- (ACC) transports, where the wind energy instead increases ual upwelling is designed to measure the vertical transfer of eddy kinetic energy at the mesoscale (e.g. Hallberg and dense waters to the upper ocean, and thus is a measure of Gnanadesikan 2001, 2006; Meredith and Hogg 2006); this process is referred to as eddy saturation. An increase in wind forcing is also expected to increase the Eulerian-mean overturning, which may be compensated by an increase in the opposing eddy-driven overturning (Böning et al. 2008; Viebahn and Eden 2010; Abernathey et al. 2011; Wolfe and Cessi 2011); this process is referred to as eddy compen- sation. Eddy saturation and eddy compensation need not always be tightly coupled. For example, some high-reso- lution model integrations reveal the eddy saturated limit is approached, but not the eddy compensation limit, so that increasing Southern Ocean winds may enhance the rate of overturning without large changes in ACC transports (Mer- edith et al. 2012; Morrison and Hogg 2013; Munday et al. 2013). Furthermore, the response of Southern Ocean large- scale circulation and its density structure to changes in physical forcing is likely to occur on centennial to millen- nial timescales. Consequently, any changes in circulation are unlikely to be detected in the relatively short available time series of observations or high-resolution model inte- Fig. 1 A schematic view of the residual upwelling connecting the grations (Allison et al. 2011; Jones et al. 2011; Shakespeare deep ocean to the atmosphere. The residual upwelling involves both and Hogg 2012). The lack of a unified view of how physi- an upper cell, called the residual circulation, including the com- cal forcing affects atmospheric CO on longer timescales is bined effect of a northward Ekman upper transport and a southward 2 eddy upper transport, and a lower cell, including sinking and north- hampering our understanding of how to place a context for ward spreading of bottom waters. There is a CO2 outgassing to the past and ongoing climate variability. atmosphere (vertical arrows) from the upwelling of carbon rich-deep Motivated by the importance of the Southern Ocean for waters and a CO2 influx into the ocean through the subduction of climate change in the present day and the past, here we pro- upper ocean waters and formation of bottom waters. The thick full arrows represent overturning streamlines and the thin lines the poten- pose a new metric—the strength of Southern Ocean resid- tial density surfaces. The subtropical upper waters (light red) are ual upwelling—to represent the dominant effect of South- warm and contain low carbon and nutrients, while the colder deeper ern Ocean dynamics on atmospheric CO2 on centennial and bottom waters (white and blue) are nutrient and carbon rich 1 3 The impact of Southern Ocean residual upwelling on atmospheric CO2 on centennial and… the communication of the deep ocean and the atmosphere 2 Model design of experiments over the Southern Ocean. This definition differs from alter- native views focusing on parts of the overturning circula- To test the skill of our metric, suites of coupled physi- tion, either the wind-induced Eulerian mean upwelling or cal and biogeochemical ocean model experiments are the residual circulation, neither of which fully account for performed with parameterized ocean eddies integrated the combined effects of the upper and lower overturning globally (Lauderdale et al.